Limitations of Using Mouse-over with Menu Item Selection

Abstract

Previous studies comparing input techniques with target acquisition tasks (where a user simply places the cursor over a designated target on the screen) found the mouse-over technique to be faster than point-and-click for targets of varying widths (Bohan & Chaparro, 1998, Bohan, Chaparro, and Scarlett 1998). This study investigated the use of the mouse-over and point-and-click techniques for a menu item selection task. Menu item selection is a more complex task than simple target acquisition because the user must read the menu text in order to select the appropriate item. An example of the use of such a task with the mouse-over technique is the Windows'98 “Start Menu”. In this study users were asked to find particular items in a cascading menu structure that consisted of a custom 4 times 3 hierarchy of common objects or locations, using either a mouse-over or point-and-click technique. The mouse-over technique required users to position the mouse over a menu item to expand the cascading menus. The user then clicked on a desired item in the third menu to select it. The point-and-click technique required the user to click the mouse on a menu item to expand the cascading menus and to click on an item on the third menu to select it. Mouse-over duration was set to 500 ms, which is comparable to the default duration for the Windows'98 Start menu. Results indicate that performance with the mouse-over technique took significantly longer than the traditional point-and-click technique (M = 5.25 seconds and M = 4.35 seconds respectively). In addition, error rates were greater with the mouse-over technique than with the point-and-click (4.95% and 2.60% respectively). Subjective data from the test participants indicate that they preferred the point-and-click technique to the mouse-over technique because they “felt more in control”. In addition, they found the “flashing menus” as they traversed the menus to the desired item using the mouse-over technique to be “distracting”. This is an interesting finding given that in target acquisition tasks, the mouse-over technique was found to be faster than the point-and-click technique. Designers need to carefully consider the tasks users will do with an application when considering the types of user interface interaction. These results show that mouse-over interaction may not be optimal in terms of task time, accuracy, and user satisfaction for menu item selection.